Finding the sources of missing heritability in a yeast cross

Abstract

For many traits, including susceptibility to common diseases in humans, causal loci uncovered by genetic-mapping studies explain only a minority of the heritable contribution to trait variation. Multiple explanations for this 'missing heritability' have been proposed. Here we use a large cross between two yeast strains to accurately estimate different sources of heritable variation for 46 quantitative traits, and to detect underlying loci with high statistical power. We find that the detected loci explain nearly the entire additive contribution to heritable variation for the traits studied. We also show that the contribution to heritability of gene-gene interactions varies among traits, from near zero to approximately 50 per cent. Detected two-locus interactions explain only a minority of this contribution. These results substantially advance our understanding of the missing heritability problem and have important implications for future studies of complex and quantitative traits.

Figure 1. Heritability for 46 yeast traits

The broad-sense heritability (H²) for each trait (X-axis) is plotted against the narrow-sense heritability (h²; Y-axis). Error bars show standard errors in heritability estimates. The diagonal line represents h² = H² and is shown as a visual guide.

Figure 2. Most additive heritability is explained by detected QTL

(A) The narrow-sense heritability (h²) for each trait (X-axis) is plotted against the total variance explained by detected QTL (Y-axis). Error bars show standard errors. The diagonal line represents (variance explained by detected QTL) = h² and is shown as a visual guide. (B) The average fraction of additive genetic variance explained (Y-axis) is plotted against number of segregants used for QTL detection (X-axis). Error bars show standard errors. Alternating shaded bands denote chromosome boundaries.

Figure 3. QTL detection for a complex trait

LOD score (log₁₀ of the odds ratio for linkage) is plotted against the genetic map. Red stars indicate statistically significant QTL. (A) LOD score plot with 1005 segregants for growth in E6-berbamine. (B) LOD score plot with 100 segregants for growth in E6-berbamine. The 15 significant QTL in A explain 78% of the narrow-sense heritability, compared with 21% for the 2 significant QTL in B.

Figure 4. Prediction of segregant trait values from QTL phenotypes

The observed phenotypic values for growth in lithium chloride (Y-axis) are plotted against the predicted phenotypic values based on a cross-validated additive model of 22 QTL. The additive QTL model explains 88% of the narrow-sense heritability. The diagonal line represents (observed phenotype) = (predicted phenotype) and is shown as a visual guide.

Figure 5. Non-additive genetic variance explained by QTL-QTL interactions

A histogram of the fraction of non-additive genetic variance explained by detected QTL-QTL interactions per trait is plotted. The histogram is restricted to traits for which at least 10% of the total genetic variation is non-additive. (Inset) Phenotypes for growth in maltose are shown, grouped by two-locus genotypes at the two interacting QTL on chromosomes VII and XI. This QTL-QTL interaction explained 71% of the difference between broad-sense and narrow-sense heritability.